Robert Fromont, Jennifer Hay
LaBB-CAT is an open-source tool for storing and automatically annotating speech corpus transcripts, using an annotation graph model for flexible annotation.
The previous ONZE Miner tool could store and annotate Transcriber transcripts but lacked support for annotations at multiple granularity levels and complex annotation structures due to its data model limitations.
ONZE Miner was redeveloped to use annotation graphs as its data model, enabling greater flexibility for automatic and manual annotation at various independent levels of granularity and supporting more sophisticated annotation structures.
LaBB-CAT opens up new possibilities for corpus mining and conversion between tool formats. Presented at the Australasian Language Technology Association Workshop in 2012.